Here’s a list of the major projects I’ve worked on and my key publications. You might also be interested in my presentations. My CV contains a compact listing of all my publications, plus any other career details you care to read. My ORCID profile also has an up-to-date list of publications.
I currently work with Joel Greenhouse to design statistical models to predict crime by using crime hotspots, spatial features, seasonal factors, and leading indicators (like 311 calls, criminal mischief, and so on). My goal is both to improve crime prediction and to provide inference tools for criminologists to understand factors that lead to crime. I also work on evaluation and diagnostic methods to understand the performance of predictive policing models.
This work is supported by a National Institute of Justice Graduate Research Fellowship (GRF-STEM).
As an undergraduate I started a project (supervised by Alex Athey of Applied Research Laboratories) to devise methods to continuously monitor the radiation background in a wide area and detect any sudden changes, such as might be introduced by a dirty bomb or stolen radioactive source. We built a system which uses gamma spectroscopy to compare new measurements to previous observations of the radiation background, making it feasible to monitor a wide area with mobile detectors and rapidly detect changes.
At Carnegie Mellon University, I continued the project under Valérie Ventura and Chad Schafer, proposing a new method based on Kolmogorov–Smirnov tests. James Scott and Wesley Tansey continued the work to devise a new spatial smoother for radiation spectra.
I have an active interest in statistical pedagogy, and in developing new ways to improve student learning, assess understanding of statistical concepts, and better teach the foundations of statistical reasoning. I’m part of the Teaching Statistics Group at Carnegie Mellon University’s Department of Statistics & Data Science.
My book Statistics Done Wrong was published in March 2015 by No Starch Press. Covering common statistical errors present in a wide range of current scientific research—not just simple misuses of t tests but failures of statistical power, unnoticed pseudoreplication, effect size inflation, and much more—it has since been translated into German, Chinese, Japanese, and Korean.
Other articles in the popular press: